Document downloaded from:

This paper must be cited as:

The final publication is available at

Copyright

http://dx.doi.org/10.1016/j.actbio.2013.07.038

http://hdl.handle.net/10251/47608

Elsevier

Vallés Lluch, A.; Arnal Pastor, MP.; Martínez Ramos, C.; Vilariño Feltrer, G.; Vikingsson,L.; Castells Sala, C.; Semino, CE.... (2013). Combining self-assembling peptide gels withthree-dimensional elastomer scaffolds. Acta Biomaterialia. 9(12):9451-9460.doi:10.1016/j.actbio.2013.07.038.

Elsevier Editorial System(tm) for Acta Biomaterialia Manuscript Draft Manuscript Number: AB-13-384R2 Title: COMBINING SELF-ASSEMBLING PEPTIDE GELS WITH 3D ELASTOMER SCAFFOLDS Article Type: Full Length Article Keywords: scaffold; poly(ethyl acrylate); self-assembling peptide gel; controlled release; cell seeding Corresponding Author: Prof. Manuel Monleón Pradas, PhD Corresponding Author's Institution: Universidad Politécnica de Valencia First Author: Ana Vallés Lluch, Ph.D. Order of Authors: Ana Vallés Lluch, Ph.D.; María Arnal-Pastor, MSc; Cristina Martínez-Ramos, PhD; Guillermo Vilariño-Feltrer, MSc; Line Vikingsson, MSc; Cristina Castells-Sala, Msc; Carlos Eduardo Semino, PhD; Manuel Monleón Pradas, PhD Abstract: Some of the problems raised by the combination of porous scaffolds and self-assembling peptide (SAP) gels as constructs for tissue engineering applications are for the first time addressed. Scaffolds of poly(ethyl acrylate) and the SAP gel RAD16-I were employed. The in situ gelation of the SAP gel inside the pores of the scaffolds was studied. The scaffold-cum-gel constructs were characterized morphologically, physico-chemically and mechanically. The possibility of incorporating an active molecule (bovine serum albumin, taken here as a model molecule for others) in the gel within the scaffold's pores was assessed, and the kinetics of its release in PBS was followed. Cell seeding and colonization of these constructs was preliminary studied with L929 fibroblasts and checked afterwards with sheep adipose-tissue derived stem cells (ASCs) intended for further preclinical studies. Static (conventional) and dynamically assisted seedings were compared for bare scaffolds and the scaffold-cum-gel constructs. The SAP gel inside the pores of the scaffold significantly improved the uniformity and density of cell colonization of the 3D structure. These constructs could be of use in different advanced tissue engineering applications where, apart from a cell-friendly ECM-like aqueous environment, a larger-scale three-dimensional structure able to keep the cells in a specific place, give mechanical support and/or conduct spatially the tissue growth could be required.

1

COMBINING SELF-ASSEMBLING PEPTIDE GELS

WITH 3D ELASTOMER SCAFFOLDS

A. Vallés-Lluch1, M. Arnal-Pastor

1*, C. Martínez-Ramos

1*, G. Vilariño-Feltrer

1, L.

Vikingsson1, C. Castells-Sala

2, C.E. Semino

2, M. Monleón Pradas

1,3,**

1Center for Biomaterials and Tissue Engineering, Universitat Politècnica de València,

Cno. de Vera s/n, 46022, Valencia, Spain

2Departament de Bioenginyeria IQS, Universitat Ramon Llull, Via Augusta 390, E-

08017 Barcelona, Spain

3Networking Research Center on Bioengineering, Biomaterials and Nanomedicine

(CIBER-BBN), Spain

*Equal contribution

**Corresponding author. Tel.: +34963877277; fax: +34963877276. E-mail:

mmonleon@ter.upv.es

Abstract

Some of the problems raised by the combination of porous scaffolds and self-

assembling peptide (SAP) gels as constructs for tissue engineering applications are for

the first time addressed. Scaffolds of poly(ethyl acrylate) and the SAP gel RAD16-I

were employed. The in situ gelation of the SAP gel inside the pores of the scaffolds was

studied. The scaffold-cum-gel constructs were characterized morphologically, physico-

chemically and mechanically. The possibility of incorporating an active molecule

(bovine serum albumin, taken here as a model molecule for others) in the gel within the

scaffold’s pores was assessed, and the kinetics of its release in PBS was followed. Cell

*Text onlyClick here to view linked References

2

seeding and colonization of these constructs was preliminary studied with L929

fibroblasts and checked afterwards with sheep adipose-tissue derived stem cells (ASCs)

intended for further preclinical studies. Static (conventional) and dynamically assisted

seedings were compared for bare scaffolds and the scaffold-cum-gel constructs. The

SAP gel inside the pores of the scaffold significantly improved the uniformity and

density of cell colonization of the 3D structure. These constructs could be of use in

different advanced tissue engineering applications where, apart from a cell-friendly

ECM-like aqueous environment, a larger-scale three-dimensional structure able to keep

the cells in a specific place, give mechanical support and/or conduct spatially the tissue

growth could be required.

Keywords: scaffold, poly(ethyl acrylate), self-assembling peptide gel, controlled release,

cell seeding

3

1. Introduction

Regenerative purposes impose multiple requirements to the material constructs intended

to help those aims. In cardiac tissue engineering, for instance, the material vehicle for an

efficient cell supply should ideally, among other things, host a large number of cells,

guarantee their viability in situ, allow a rapid vascularization, and maybe give

mechanical support to the infarcted myocardial tissue; this last requirement, in the case

of neural tissue engineering, must be replaced by the appropriate guidance of axonal

growth. It would be unlikely that the multifunctionality demanded for tissue engineering

constructs by these applications could be achieved by simple, one-component material

structures. Here we explore to which extent the combination of two different materials

imparts to the resulting construct definite advantages from a tissue engineering point of

view.

Synthetic or natural hydrogels possess many properties similar to those of the

extracellular matrix (ECM), which makes them good candidates for engineered cellular

niches. However, these gels cannot meet demanding mechanical requirements, which

limit greatly their uses. Mechanical strength and geometrical definiteness should be

provided by porous scaffolds. In a previous work [1] we studied how to incorporate a

hyaluronic acid hydrogel into the porous structure of poly(ethyl acrylate) (PEA)

elastomer scaffolds. Problems related to the filling technique, to the in situ crosslinking

of the gel, and to the degree of pore coating or filling depending on the gel

concentration, had to be addressed there. Other scaffolds with internal gel coating, such

as the couples chitosan/fibrin [2], or polylactide/HA [3], have been studied; in each

case, the filling and the gelation or crosslinking inside the pores represented specific

difficulties to overcome. In the present paper we study constructs consisting of PEA

4

scaffolds with the self-assembling peptide (SAP) RAD16-I hydrogel filling the pores.

Some general questions raised by the combination of these very different materials are

addressed here. How should the scaffold be filled? Can the peptide solution within the

pores be gelled in situ? How should the cells be incorporated into the scaffold? Will the

cells diffuse through the SAP lodged previously in the scaffold? Does indeed the

construct show better biological performance than the empty scaffold?

Self-assembling peptide (SAP) hydrogels are a family of promising synthetic materials.

They are injectable in aqueous solution and undergo spontaneous assembling into

ordered nanostructures [4,5]. In this work, the RAD16-I peptide (commercially

available as BDTM

PuraMatrixTM

) has been used; it consists of a sequence of 16

aminoacids AcN-(RADA)4-CONH2 (R: arginine, A: alanine and D: aspartic acid),

which forms β–sheet nanofiber-like gels as a response to pH or ion concentration

changes [4,6-,9]. In vitro it encourages the adhesion of endothelial cells, their

proliferation and formation of capillaries [10,11] to enhance cardiac myocyte survival

[12], and to support the survival, differentiation capacity and migration of neural stem

cells [13] and the formation of synapses [14]. In mice, when injected with cardiac

progenitors, it promotes angiogenesis and differentiation into cardiomyocytes and

vascular smooth muscle cells [15]. Furthermore, as a wholly synthetic ECM-like

medium, it presents the advantages, over other protein gels, of being easily

functionalizable and of being free from pathogenicity worries. However, this SAP gel is

but a little more consistent than an acqueous solution [16]; it does not meet the demands

of shape stability and mechanical resistance of many tissue engineering applications.

Thence the interest of its possible combination with other structurally more consistent

materials. In [17] a RAD16-I solution containing Schwann cells was employed to fill

5

the lumen of 2.2 mm-diameter cellulose channels, to study regeneration across 10 mm

gaps after peripheral nerve injury.

PEA is, at body temperature, a rubber when cross-linked, and its mechanical properties

(elastic and loss moduli) have values not dissimilar to those of soft biological tissues. It

has shown excellent compatibility with different types of cells in vitro: chondrocytes

[18], keratinocytes [19], endothelial cells [20], neural cells [21-24], osteoblasts [25] or

dental pulp stem cells [26], and has been recently proposed as a feeder-free platform for

the maintenance and growth of human embryonic stem cells [27]. Due to its easy way

of polymerization, PEA can be prepared in the form of scaffolds with different

architectures, liable to invasion by cells and intended as guiding support for cell growth

[1,19,20,22,23,28-31].

Here we establish a protocol for the assembly of the RAD16-I peptide gel within the

pores, having diameters smaller than a hundred microns, of hydrophobic PEA scaffolds,

and characterize the performance of the combination as delivery platforms and as cell

growth niches. The present is, up to our knowledge, the first work that studies such

system hydrophobic scaffold plus RAD16-I hydrogel as in situ assembling filler of the

micropores. Though these constructs are intended for future tissue engineering

applications, the first studies here presented, due to their preliminary methodological

nature, were performed using a commercial line of mouse fibroblasts; they have served

to choose the main assembly conditions and to understand the role of the gel as a cell

diffusion and encapsulation medium. The system here presented with ASCs is currently

being employed in animal experiments as cell carrier to support the regeneration of

cardiac tissue after myocardial infarction. In the present paper we characterize the

assembly and cell seeding of these constructs.

6

2. Materials and methods

2.1. Preparation of the bare elastomer scaffolds

Poly(ethyl acrylate) scaffolds with interconnected spherical pores were obtained by a

radical polymerization of the monomer mixture and a template leaching technique,

following a procedure analogous to that used in [29]. Briefly, ethyl acrylate (99%,

Aldrich) monomer was mixed with 1 wt% of benzoin (98%, Scharlau) as photo-initiator

and 2 wt% of ethyleneglycol dimethacrylate (98%, Aldrich) as cross-linker, stirred for

15 min, injected in the corresponding porogen template, polymerized for 24 h under a

UV source, and post-polymerized for 24 h more in an oven at 90ºC. 5 x 5 cm2 templates

were previously obtained by sintering poly(methyl methacrylate) microspheres of 90 ±

10 μm in diameter (PMMA; Colacryl dp 300). After polymerization, the PMMA

templates were removed by soxhlet extraction with acetone (Scharlab). After this, the

solvent was slowly exchanged with water, and the scaffolds were dried under vacuum

and stored until use. Their final thickness was 0.8 mm approximately.

Bulk 0.8 mm-thick PEA sheets (hereafter 2DPEA) were synthesized by an analogous

procedure, to be used as two-dimensional controls. The monomeric mixture was

allowed to polymerize in a mold consisting of two glass plates with a rubber separator

in between. Once polymerized, the sheets were rinsed in boiling acetone for 48 h with a

solvent renewal and finally dried.

Samples for the different experiments were cut from the scaffolds or the bulk sheets as 8

mm-diameter circles. Bulk samples were employed to determine the water contact angle

and swelling of the PEA matrix, because in its porous format the pores would interfere

with such measurements.

7

2.2. Preparation of the scaffold-cum-gel constructs

The SAP hydrogel RAD16-I (PuraMatrix™ 1% (w/v), BD Biosciences) was used as a

filler of the scaffolds’ pores. In order to decrease the viscosity of the stock solution, the

original package of RAD16-I was placed in a bath sonicator (Bandelin) for 30 min at

25ºC and 30 W were applied. The stock solution was then diluted with water (extra

pure, Scharlau) to a ready-to-use 0.15% (w/v) concentration (high enough to reach

percolation of its network) one and vortexed (Elmi SkyLine) to ensure

homogeneization. In order to facilitate the penetration of the still viscous SAP solution

into the micropores of the hydrophobic PEA scaffolds, three different procedures were

followed. The first method involved the incorporation of the peptide solution, loaded in

a syringe, into the scaffolds by applying some pressure. The second method consisted in

improving the wettability of the PEA scaffolds by a plasma treatment, and afterwards

soaking them with the peptide solution. A Plasma-Electronic Piccolo chamber was

employed, evacuated to 4∙10-2

mbar before generation of Ar/O2 (1:1) plasma at 25 W for

5 min. The third filling procedure consisted in the introduction of the scaffolds, together

with the SAP solution, in a tube closed with a septum rubber stopper, pierced by a

syringe’s needle, and the application of vacuum performing successive movements of

the piston.

Once filled with the peptide solution, the scaffolds were placed in a Petri dish and drops

of phosphate buffered saline (PBS) solution or culture medium were added to their

boundaries to induce the self-assembly of the peptides. The SAP was allowed to gel in

the pores of the scaffolds for 30 or 60 min, which are the limits of the time range

recommended by the manufacturer to obtain free gels. The obtained constructs were

used for measurements immediately to avoid drying. Bare PEA scaffolds were kept as

controls.

8

2.3. Morphology of the scaffold-cum-gel constructs

Bare scaffold samples were examined by scanning electron microscopy (SEM) in a JSM

6300 (JEOL Ltd., Tokyo, Japan) device, with the samples previously sputter-coated with

gold, 15 kV of acceleration voltage and 15 mm of working distance. The samples were

fractured in liquid nitrogen in order to obtain images of the surface and the cross-

sections.

The effective average pore diameter was determined with the free ImageJ software

(National Institute of Mental Health, Bethesda, Maryland, USA) from 3 cross-section

images at 200X. A set of pores were chosen and outlined, and the circularity of each one

was calculated. The average pore diameter was obtained from the areas of 30 pores

having circularities above 90%.

The effectiveness of the filling procedure and the in situ gelation of the SAP once inside

the scaffold’s pores was assessed by cryoSEM in a JSM5410 (JEOL Ltd., Tokyo, Japan)

instrument equipped with a cryounit (Oxford CT 1500). The samples were fixed

vertically in a metal sample holder and submerged in liquid nitrogen to cut off a cross

section and observe it. Water entrapped by the SAP before and after gelation was

sublimated at -70ºC in the cryogenic unit for 15 min in vacuum so that the structure it

left in the pores could be observed. Then, the samples were gold-sputtered and observed

at 15 kV and 15 mm of working distance. PBS injected in the pores following the

procedure described above was employed as filling control for cryoSEM.

In parallel, the effective gelation (with PBS) of the SAP within the pores of the scaffolds

was assessed by staining with a Congo red 0.1% (w/v) acqueous solution (Fischer

Scientific) for 20 min followed by 30 min of rinsing with water and macroscopical

observation. Scaffolds with only PBS in their pores were also stained as controls.

9

2.4. Porosity determination

The pore volume fraction of the bare scaffolds was obtained with the help of the density

of PEA (1.13 g/cm3 [29]), and the weight and apparent volume of the scaffolds, by

triplicate. The porosity, π, is defined as the pore volume fraction of the scaffolds:

pores

app app

V mv1

V V (Eq. 1)

where Vpores

is the volume of pores, Vapp

is the geometric (or apparent) volume,

calculated from the measured linear dimensions of the scaffold, m its weight and v the

PEA specific volume (reciprocal of its density). A Mettler AE 240 balance (Mettler-

Toledo Inc., Columbus, OH, USA), with a sensitivity of 0.01 mg, was employed for this

purpose. These tests were also carried out on scaffolds filled with SAPs to confirm the

complete filling of the pores.

2.5. Contact angle and swelling

The PEA wettability was determined by measuring the average contact angle of extra

pure water drops on the surface of 7 samples of 2DPEA in a Dataphysics OCA

instrument (DataPhysics Instruments GmbH, Filderstadt, Germany).

To evaluate the PEA swelling, four 50 mm2 pieces of 2DPEA sheets were weighed

vacuum dried and after different times of immersion in PBS at 37ºC until equilibrium

was confirmed, using the previously mentioned balance. The equilibrium water content,

EWC, is defined as the ratio of the mass of absorbed water at equilibrium to the mass of

dry material.

2.6. Mechanical properties

10

Mechanical compression tests were performed on cylindrical scaffolds having a diameter

of 8 mm, bare, filled with water and with the SAP gel, 5 replicas in each case. An

EXSTAR TMA/ss6000 equipment (Seiko Instruments Inc., Chiba, Japan) was used for

this purpose, the load increasing from 0 to 150 g at 10 g/min, at room temperature.

Before measurements, the excess of SAP solution on the external surface of the scaffolds

was carefully removed with a spatula. Measurements on empty scaffolds and on

scaffolds filled with water were carried out as references. The compressive elastic

moduli were obtained as the slope of the stress-strain plots in the linear deformation

regions, at low deformations (E1) and at strain 0.3 (E2).

2.7. BSA uptake and release from the scaffold-cum-gel construct

The SAP solution was included within the pores of three pieces of PEA scaffolds of 8

mm of diameter, and gelled with PBS drops during 30 min, as described. The scaffolds

were weighed bare and after being filled with the gel to obtain the total amount of SAPs

able to absorb the protein, mgel

, as the difference. Bovine serum albumin (BSA; 2 mg/ml

in 0.9% acqueous solution, Thermo Scientific) solution at a concentration, BSAsol

ini, of 1

mg/ml was prepared by dilution in PBS. The scaffold-cum-gel constructs were then

submerged in the BSA solution (1.6 ml per piece, Vsol

) and left overnight at 37ºC,

allowing the load of BSA into the gel. During this time, the scaffolds were held upright

in the vials with the help of a metal needle. Immediately after absorption (to avoid gel

drying), each BSA-filled scaffold-cum-gel piece was placed upright in a vial with 1.2 ml

of PBS at 37ºC to follow the BSA release from the gel.

The supernatant was collected at pre-selected time intervals (every hour during the first 8

h, and every 24 h from this moment on) up to 10 days and replaced with fresh PBS

tempered at 37ºC. 1 ml of each supernatant, s .natantV , was allowed to react with BCA

11

reactant (Thermo Scientific) in a 1:1 volume ratio for 1 h at 60ºC. The mixtures, giving a

purple colour with absorbance at 562 nm proportional to the BSA concentration, were

scanned in an ultraviolet spectroscope (Cecil CE9200, Aquarius). A standard calibration

curve was obtained with solutions of known concentrations of BSA, [BSA], up to 50

μg/ml. This standard curve allowed to correlate the absorbance at each time with the

mass of BSA released in each period, making use of the supernatant

volume: BSA s .natantV .

In parallel, the BSA solution exhausted after the overnight load was diluted as needed

and analysed as described above by UV spectrophotometry and with the help of the

calibration curve to determine the residual concentration, BSAsol

res. The amount of BSA

initially loaded in the gel within the pores of the scaffolds, M0, was calculated as the

difference between the initial BSA mass in the loading solution and the mass remaining

in the solution after sorption overnight, BSA BSAsol sol sol

0 ini resM V . The

cumulative mass released up to time t, Mt, was divided by M0 to obtain the percentage of

BSA released at each time.

2.8. Cell-seeding issues and cell cultures in the scaffold-cum-gel constructs

Commercial L929 mouse fibroblasts from subcutaneous connective tissue, areolar and

adipose (Sigma Aldrich) in their 10th

passage were employed for preliminary in vitro

tests. The bare scaffolds were sterilized previously with a 25 kGy dose of gamma

radiation from a 60

Co source (Aragogamma, Barcelona, Spain). L929 cells were seeded

in two series of materials: scaffolds whose pores had been filled with a non-gelled 0.15

% (w/v) SAP solution with the help of some vacuum were compared with scaffolds

whose pores had been previously filled with PBS. Two seeding methodologies were

12

compared. (a) In the one hereafter called “static seeding”, 105 cells were seeded on the

upper-surface of each piece of scaffold having a diameter of 8 mm in a 48-well tissue

plate, suspended in 20 μl of a 10% sucrose (Sigma-Aldrich) acqueous solution. The

samples were kept for 30 min in the incubator at 37ºC, and then 380 μl of culture

medium consisting in DMEM high glucose supplemented with 10% FBS, 1%

penicillin-streptomycin and 1% L-glutamine (Sigma) were added to each well. (b) In the

“dynamic seeding” methodology, the cells were seeded on the scaffolds in the same

number and way as above, but then the culture plate was smoothly shaken in a Titramax

101 shaker (Heidolph instruments, Germany) for 30 min; then, 380 μl of culture

medium were added to each well. The scaffolds seeded in both ways were cultured at

37°C in a humidified atmosphere under 5% CO2 for 1, 3 and 7 days. The culture

medium was renewed every 2-3 days.

In order to confirm that the assembly conditions established with the L929 cell line

were effective also with the cells employed in a big animal model of myocardial

infarction the procedures were repeated with adipose-tissue derived stem cells (ASCs).

ASCs were provided by the Hôpital Georges Pompidou (Paris); they were isolated from

fat biopsies of the right thoracic wall of female Rambouillet sheeps as described in [32].

ASCs were seeded in a 20 μl sucrose-enriched aqueous droplet after the loading of the

peptide. After a smooth shaking, 380 μl of culture medium was added and the culture

was prolonged for 7 days.

2.9. Analysis of the cell distribution in the scaffold-cum-gel constructs

After culture, cell proliferation and distribution were observed by fluorescence (Leica

DM6000) or confocal laser scanning microscopy (Olympus, FV1000). Previously, the

samples were rinsed with PBS, fixed with 4 % paraformaldehyde (PFA; Panreac) for 20

13

min at room temperature, and washed twice with PBS for 5 min. Cells were

permeabilized with 0.1 vol% Triton X-100 for 30 min at room temperature. After two

rinses with DPBS, the samples were incubated with BIODIPY-FL phallacidin

(Invitrogen, 1:200) and vimentin (Sigma, 1:100) for 1 h at room temperature in the dark,

next washed twice with DPBS and counterstained for 5 min with 4′,6-diamidino-2-

phenylindole dihydrochloride (DAPI; Sigma, 1/5000). The pieces were rinsed twice

with PBS, cryopreserved in 30% sacarose and finally cut in 100 μm slices. For

microscopy observation, the slices were mounted with Fluorsave reagent (Calbiochem)

coverslipped, and stored in the dark at 4°C until use.

The morphology of the cells was observed by scanning electron microscopy (SEM) in a

JSM-5410 microscope. For this, samples were fixed with 3.5% glutaraldehyde (GTA;

Scharlab) for 60 min at 37ºC, followed by two rinsings with PBS. Next, samples were

incubated with 2% osmium tetraoxide for 2 h at room temperature followed by four

rinses with distilled water. Afterwards, samples were dehydrated through a series of

graded ethanol (30, 50, 70, 96, and 100) at 4ºC. Samples were finally dried by critical-

point drying and sputter-coated with gold to be observed by SEM at 15 kV and 15 mm

of working distance.

3. Results

3.1. Morphology and physical properties of the scaffold-cum-gel constructs

PEA scaffolds obtained by the inverse template technique have interconnected spherical

pores, which leave trabeculae as they intersect (Figure 1). The measured pore volume

fraction of the structure is 80.8±3.5%, and the average pore diameter of 90.96 ± 6.91

µm. The equilibrium water content of 2DPEA, attained after 3 days of immersion in

14

PBS, resulted in 1.14 ± 0.16%. The water contact angle on 2DPEA is characteristic of a

hydrophobic difficult to wet material, 72.9 ± 4.1º.

Natural diffusion was not enough to fill the scaffolds with the peptide solution due to

their hydrophobicity, and alternative loading procedures were proposed. The application

of pressure to force the solution to penetrate into the pores yielded heterogeneously

filled scaffolds, even after a subsequent centrifugation. The plasma-assisted procedure,

though effective to hydrophilize the surface of the scaffolds, does not seem to be a

viable strategy in the clinical practice. For these reasons, these methods were not further

considered. The third procedure described above to incorporate the SAP consisted in the

introduction of the scaffold in a syringe with the peptide solution and helping the

intrusion of the solution by evacuation. This method allowed to completely fill the

scaffold pores with the SAP solution. The solution could be brought to gelation in situ

within the pores in 30 minutes. No differences were observed with those samples gelled

for longer times; thus, 30 min was hereafter the gelation time employed. When water

sublimates for cryoSEM observation, the trace of the non-gelled SAP solution appears

as parallel walls (Figure 2 b); once gelled, the SAP nanofibrilar network leaves a

honeycomb-like structure in the pores of the scaffolds (Figure 2 c and d). These

structures cannot be observed after water sublimation in the scaffold’s pores containing

PBS (Figure 2 a). Macroscopically, Congo red stains the gel within the pores in red,

demonstrating the formation of a β-sheet structure, in contrast with the scaffolds

containing only PBS, which remain white (insets of Figure 2 a and c).

The compression curves of the scaffolds, either bare, filled with water or filled with the

RAD16-I solution and gelled with PBS, show an overall non-linear dependence of stress

upon strain with increasing modulus, Figure 3. A first linear deformation region ranges

up to 5% strain and 10 kPa stress, followed by a region of increasing slope, which

15

stabilizes in a second linear deformation region for strains greater than 25% at stresses

above 100 kPa. The Young moduli in both linear deformation regimes are included as

insets in Figure 3. Those of the scaffolds filled with water were found to be 0.053±0.01

MPa and 0.63±0.15 MPa, respectively. The moduli of the scaffolds filled with the gelled

peptide are slightly greater than those of bare or filled with water ones.

3.2. BSA uptake and release from the scaffold-cum-gel constructs

The concentration of BSA incorporated to the scaffold’s pores, obtained from the mass

uptake M0 and the volume of SAP solution in the pores, resulted in 12.7 mg/ml.

Unexpectedly, this value is much higher than that of the starting BSA solution, 1 mg/ml.

Figure 4 shows the fractional BSA, Mt/M0, release curve of the loaded construct, where

M0 is the initial mass of BSA in the scaffold-cum-gel construct, and Mt is the mass

released from it at time t. Two different release regimes can be identified: first a fast

protein release up to 14.9% of the initial amount of BSA within the first eight hours, and

then a slower one from this moment on up to 25% in 10 days. The initial release regime

can be successfully fitted to the power law Mt/M0 = 0.0785*t0.33

, with time t in hours; the

second release one is almost linear in time and follows Mt/M0 = 0.0004*t+0.1493, for t

greater than 8 h.

3.3. Cell-seeding and cell cultures in the scaffolds and scaffold-cum-gel constructs

Bare scaffolds filled with PBS and scaffolds filled with the SAP solution were seeded

on their upper surface with fibroblasts in an aqueous droplet with sucrose. In previous

tests cells were seeded with culture medium, but this provoked the gelation of the

peptide as cells started to diffuse through it, which impeded a homogeneous

colonization of the scaffold.

16

After seeding, cells were allowed to diffuse within the scaffold with or without

mechanical assistance, and after 30 min culture medium was added to induce gelation.

Figures 5 to 7 show that more cells colonize the scaffolds when the scaffold’s pores are

filled with the SAP gel, and consequently after 7 days they have proliferated and occupy

all the pores. Comparison of the images obtained from scaffolds seeded statically and

dynamically clearly indicates that the dynamic seeding favors the invasibility of the

pores by the cells, which remain encapsulated when the peptide fills the pores,

proliferate and are much more numerous after 7 days. Dynamic cell seeding for

scaffolds containing the SAP gel leads after 1 day to a uniform distribution of the cells

throughout the scaffold, Figure 7. The cells seem to be leaning on the polymeric

trabeculae at this culture time (see the arrowheads pointing at cells around a spherical

pore in Fig. 7), and at 7 days completely invade the lumen of the pores (the area

occupied by the scaffold is clearly seen in dark -see the asterisks-, delimiting the stained

cells in the pores). At this culture time the cells display under SEM either an elongated

or a more rounded morphology (Figure 8 a, see the broken line outlining a pore, and the

asterisk, respectively), depending on whether they adhere to the PEA trabeculae or

swim within the pores, and they have synthesized a large amount of extracellular

matrix. More closely (Figure 8 b), such rounded cells seem to be embedded in what is

probably ECM and SAP gel completely filling the pores of the scaffold.

Sheep ASCs intended for preclinical studies were also able to invade uniformly the

whole scaffold when they were seeded after the SAP pre-load, despite their larger

dimensions (Figure 9). After 7 days, the scaffolds showed a large and homogeneously

distributed population of cells occupying completely the free spaces and even covering

both faces of the scaffold.

17

4. Discussion

The results of our study show that it is feasible to combine SAP hydrogels and

elastomeric hydrophobic scaffolds into constructs having some of the distinctive

advantages of both kinds of materials, thus producing vehicles for cell seeding and cell

supply with unique properties: a cell-friendly, ECM-like immediate microenvironment

for the cells that facilitates an efficient and uniform cell colonization of the scaffolds,

plus shape stability, mechanical consistency and strength. Furthermore, these constructs

do also possess interesting features as active molecules delivery platforms. In the

following we discuss briefly these aspects.

The concentration of SAP in the prepared solutions (0.15% w/v) for injection in the

scaffolds’ pores is so low that it essentially has the density and viscosity of liquid water;

after gelling, the viscosity is somewhat greater, but still the SAP hydrogel has the

rheological properties of a mucus, and thus negligible resistance to shearing and tensile

stresses, and to manipulation in general. The scaffold-cum-gel construct represents, in

this mechanical respect, an improvement of several orders of magnitude: from moduli

values typical of a mucus (10 to 80 Pa) we pass to values typical of an elastomer (of the

order of 1 MPa) (Figure 3), and obtain a manipulable construct, able to withstand

tractions, compressions and shears to the extent that a typical elastomer scaffold does.

The presence in the scaffold’s pores of the filler gel yields a somewhat higher modulus

in compression. This effect can be attributed mainly to the presence of water in the

pores, which is very incompressible, and to its relative difficulty to flow out of the pores

upon compression; the gelled state of the SAP results in a somewhat more hindered flow,

which translates into somewhat higher linear moduli. This phenomenon is known as

poroelasticity, and it is typical of fluid-filled porous media under compression [33,34]. In

18

our experiments the mechanical effect of the filling gel is manifest already at low strains,

and persists in the linear deformation region.

The scaffold-cum-gel construct is useful also as a platform for drug and molecule

delivery, independently of the presence of cells. BSA was employed in this study as a

model protein for uptake and release. The SAP hydrogel can be considered as a two-

phase system, one phase being a water solution of non-assembled SAP molecules, the

other phase constituted by the assembled SAP molecules into a nanofibrilar network.

The BSA molecules sorbed in the scaffold-cum-gel construct may be thought to reside

in three different placements: either solved in the bulk of the watery phase inside the

pores, or located in the assembled phase, or, finally, adsorbed onto the surface of the

scaffold’s pores. The initial concentration of BSA in the construct was estimated by

means of M0, the mass of gel entrapped in the scaffold, mgel

, and assuming that the

density of the gel is close to that of water: BSAconstruct gel gel gel

0 0iniM V M m .

This calculus resulted in a value of 12.7 mg/ml, which is much higher than that of the

starting solution from which the uptake took place, 1 mg/ml. Firstly, this finding

confirms that the protein can be easily and efficiently entrapped after the formation of

the gel. Secondly, it indicates that the construct is able to concentrate the BSA solution

(which might be another useful property for some applications). Since, by

thermodynamic equilibrium, the concentration of BSA molecules able to diffuse freely

must be similar in the gel phase inside the pores and in the outside watery phase, this

difference is indicative of BSA molecules in the construct residing preferentially either

adsorbed on the PEA surfaces or trapped within the microphase constituted by the

assembled microfibrils. Our experiments are insufficient to establish the relative

importance of both entrapment modes, and further characterization would be in need on

this point. The BSA molecule may indeed interact electrostatically with the peptide

19

chains; the amide and acid groups of its molecular structure yield a different net charge

of the BSA molecule at different pH values, its isoelectric point being 4.2 [35]. During

the BSA uptake process, which takes place above this pH, the molecule is negatively

charged and thus may interact with residual positively charged groups of the SAP

macromolecules. A supporting result is found in [36]: with increasing pH, the ionization

of the acid groups of the protein and the poly(acrylic acid) in which it was entrapped

increased the repulsion forces between them, and consequently reduced the free space

available for the protein molecules in the gel network.

The BSA loaded constructs are able to release the protein when immersed in PBS. The

kinetics of this release has two clearly differentiated time regimes, Figure 4. A faster

release kinetics occurs within the first eight hours, time at which about a 15% of the

total BSA amount is released; from that time on a much slower kinetics is in act. The

residual BSA remaining after 10 days was still 75% of the initial amount, and the plot

didn’t show yet any decrease of the time rate of release. A good fit of the first stage of

the time-release curve to a power law was obtained, Mt/M∞=k·tn (inset in Figure 4);

least-squares fitted values of the k and n parameters turned out to be 0.078 and 0.33,

respectively. A value of n = 0.5 corresponds to Fickian diffusion from a slab; that is, the

case in which the release is controlled by a pure diffusion process, a concentration

gradient equalization due to random brownian motion of the molecules, with no bias in

their displacements (Fickian diffusion from a cylinder has n = 0.45, and from a sphere n

= 0.43) [37]. This is here not the case. The fitting of the power law to the second release

regime gives n = 1 (linear growth with time of the amount released) and k = 0.0004

(plus the new parameter = 0.1493 needed to account for the initial release). Power

laws such as the stated one are typically used to describe solvent release from hydrogels

[38-40]; the parameter k is then interpreted as a structural/geometric factor to account

20

for the different tortuosities of the transport path. In [38], the authors followed the

release of BSA from previously dried poly(N-isopropylacrylamide) hydrogels, which

lasted barely 2 h at 40ºC, and obtained higher values for the k and n parameters of 0.112

and 0.462, respectively. Other BSA-gel systems have also been found to exhibit double

kinetics of BSA release, which have been interpreted as an indication of the interference

with diffusion-controlled transport of further mechanisms: swelling or shrinking of the

gel matrix (sometimes triggered by pH or temperature changes in the environment)

[38,39,41], chemical reactions (polymer degradation, drug-polymer interaction, etc.)

[36,39,42], and processing (heterogeneous cross-linking of the gel, entrapment of the

drug on the surface, use of solvents afterwards dried, etc.) or storage conditions (drug

saturation at the surface on gel drying, drug migration, etc.) leading to an uneven drug

distribution [43]. When hydrogel samples are dried after the protein loading, their

swelling usually provokes a burst release [36,38], due to those molecules near the

polymer surface and rapidly transferred to the medium. This is obviously not the case

here, as the constructs were not dried after the protein loading. The rapid initial kinetics

observed in our system could be related to the electrostatic interactions between the

BSA molecules and the oligopeptides forming the gel fibrils. As previously discussed,

the excess of BSA within the scaffold-cum-gel construct over the initial concentration

of the uptake solution must be attributed to specific interactions of BSA molecules with

the PEA polymer chains, or with the assembled SAP molecules, or with both. Thus,

release of BSA solved in the watery phase of the hydrogel may be expected to be faster

than the release of BSA molecules which are either linked to the peptide nanofibrillar

assemblies or adsorbed to the PEA surfaces. A slower release of these BSA molecules

must follow, since those specific bonds must be eased.

21

A further effect of pore filling with the SAP gel is to improve cell distribution after

seeding within the scaffolds. The presence of the (non-gelled) SAP solution within the

scaffolds’ pores improved fibroblasts colonization in terms of number of cells and

uniformity of their distribution. The SAP solution acted as a faster medium for cell

invasion of the pores of the scaffolds than mere culture medium (Figure 7, 1 day). At 7

days the cells had proliferated much more when the pores were filled with the SAP gel

(Figure 7, 7 days, and Figure 8). The dynamic seeding achieved better results than the

conventional one; 30 min of smooth shaking were enough to facilitate cell diffusion

throughout the scaffolds and a uniform distribution. The advantage of the SAP filling

for cell colonization may be due to the higher diffusion speed of the cells in this

medium than on the scaffolds surfaces: PEA is a hydrophobic polymer onto which most

cells are willing to adhere and spread, thus delaying their progress through the structure.

Swimming through the peptide solution seems a faster way of colonization than

creeping on the scaffold’s pores surfaces. In our experiments, the cells were seeded with

the minimum amount of medium, and planted as a droplet onto the SAP-filled scaffold;

this procedure avoids their spilling out of the scaffold and optimizes cell seeding. The

SAP was gelled later, consequently entrapping the cells inside the pores. This seeding

procedure was also successful with ASCs, a larger kind of cell that is currently being

employed in a preclinical study in a sheep animal model of chronic infarction; sheep

ASCs were incorporated at a high density to the SAP pre-loaded scaffolds, and allowed

to invade completely the volume occupied by the peptide solution with the help of a

smooth shaking before gelation.

The scaffold-cum-gel constructs here developed provide to the cells seeded in them a

friendly immediate ECM-like environment through the SAP gel component, while at

the same time possess enough mechanical integrity and shape stability thanks to the

22

embracing elastomeric scaffold. The ECM-like SAP gel can surround the cells,

improving the survival and migration conditions, and can, at the same time, incorporate

and deliver active molecules such as drugs or growth factors in a controlled manner to

regulate cell fate or stimulate an angiogenic response.

23

5. Conclusions

SAP solutions can be effectively incorporated and gelled within the pores of elastomeric

scaffolds. These constructs can then be seeded in a more efficient manner than bare

scaffolds, probably because the hydrogel represents a better migration medium for cells.

The scaffold-cum-gel platform can also be employed as a protein delivery vehicle with

unique features. In summary, these constructs combine several advantages of their

separate components yielding easy-to-produce and to manipulate systems for cell and

factor delivery.

24

Acknowledgements

The authors acknowledge funding through the European Commission FP7 project

RECATABI (NMP3-SL-2009-229239), and from the Spanish Ministerio de Ciencia e

Innovación through projects MAT2011-28791-C03-02 and -03. Dr. J C Chachques

(Hôpital Européen Georges Pompidou, Paris) is thanked for providing the ASCs

employed in this study.

References

1. Arnal Pastor M, Vallés Lluch A, Keicher M, Monleón Pradas M. Coating typologies

and constrained swelling of hyaluronic acid gels within scaffold pores. J Colloid

Interf Sci 2011;361:361-369.

2. Blan NR, Birla RK. Design and fabrication of heart muscle using scaffold based

tissue engineering. J Biomed Mater Res 2008;86A:195-208.

3. Antunes JC, Oliveira JM, Reis RL, Soria JM, Gómez-Ribelles JL, Mano JF. Novel

poly(L-lactic acid)/hyaluronic acid macroporous hybrid scaffolds: characterization

and assessment of cytotoxicity. J Biomed Mater Res 2010;94A:856–869.

4. Zhao X, Zhang S. Self-assembling nanopeptides become a new type of biomaterial.

Adv Polym Sci 2006;203:145-170.

5. Jung JP, Gasiorowski JZ, Collier JH. Fibrillar peptide gels in biotechnology and

biomedicine. Pept Sci 2010;94:49-59.

6. Zhang S, Gelain F, Zhao X. Designer self-assembling peptide nanofiber scaffolds for

3D tissue cell cultures. Semin Cancer Biol 2005;15:413-420.

25

7. Takei J. 3-Dimensional culture scaffold for everyone: drug screening, tissue

engineering and cancer biology. AATEX 2006;11:170-176.

8. Quintana L, Fernández Muiños T, Genové E, Olmos MM, Borrós S, Semino CE.

Early tissue patterning recreated by mouse embryonic fibroblasts in a three-

dimensional environment. Tissue Eng Part A 2008;14:1-10.

9. Ortinau S, Schmich J, Block S, Liedmann A, Jonas L, Weiss DG, et al. Effect of 3D-

scaffold formation on differentiation and survival in human neural progenitor cells.

Biomed Eng Online 2010;9:70.

10. Sieminski AL, Semino CE, Gong H, Kamm RD. Primary sequence of ionic self-

assembling peptide gels affects endothelial cell adhesion and capillary

morphogenesis. J Biomed Mater Res Part A 2008;87:494–504.

11. Genové E, Shen C, Zhang S, Semino CE. The effect of functionalized self-

assembling peptide scaffolds on human aortic endothelial cell function. Biomaterials

2005;26:3341-3351.

12. Narmoneva DA, Vukmirovic R, Davis ME, Kamm RD, Lee RT. Endothelial cells

promote cardiac myocyte survival and spatial reorganization. Circulation

2004;110:962-968.

13. Thonhoff JR, Lou DI, Jordan PM, Zhao X, Wu P. Compatibility of human fetal

neural stem cells with hydrogel biomaterials in vitro. Brain Res 2008;1187:42-51.

14. Holmes TD, Lacalle S, Su X, Liu G, Rich A, Zhang S. Extensive neurite outgrowth

and active synapse formation on self-assembling peptide scaffolds. PNAS

2000;97:6728-6733.

15. Tokunaga M, Liu ML, Nagai T, Iwanaga K, Matsuura K, Takahashi T, et al.

Implantation of cardiac progenitor cells using self-assembling peptide improves

cardiac function after myocardial infarction. J Mol Cell Card 2010;49:972-983.

26

16. Yamaoka H, Asato H, Ogasawara T, Nishizawa S, Takahashi T. Cartilage tissue

engineering using human auricular chondrocytes embedded in different hydrogel

materials. J Biomed Mater Res Part A 2006;78:1–11.

17. McGrath AM, Novikova LN, Novikov LN, Wiberg M. BD PuraMatrix™ peptide

hydrogel seeded with Schwann cells for peripheral nerve regeneration. Brain Res

Bull 2010;83;207-213.

18. Pérez Olmedilla M, García-Giralt N, Monleón Pradas M, Benito Ruiz P, Gómez

Ribelles JL, Cáceres Palou E, et al. Response of human chondrocytes to a non-

uniform distribution of hydrophilic domains on poly(ethyl acrylate-co-hydroxyethyl

methacrylate) copolymers. Biomaterials 2006;27:1003-1012.

19. Campillo-Fernández AJ, Pastor S, Abad-Collado M, Bataille L, Gómez Ribelles JL,

Meseguer Dueñas JM, et al. Future design of a new keratoprosthesis. Physical and

biological analysis of polymeric substrates for epithelial cell growth.

Biomacromolecules 2007;8:2429-2436.

20. Campillo-Fernández AJ, Unger RE, Peters K, Halstenberg S, Santos M, Salmerón

Sánchez M, et al. Analysis of the biological response of endothelial and fibroblast

cells cultured on synthetic scaffolds with various hydrophilic/hydrophobic ratios.

Influence of fibronectin adsorption and conformation. Tissue Eng 2009;15:1331-

1341.

21. Soria JM, Martínez-Ramos C, Benavent V, Campillo-Fernández A, Gómez-Ribelles

JL, Salmerón-Sánchez M, et al. Survival and differentiation of embryonic neural

explants onto different biomaterials. J Biomed Mater Res Part A 2006;79:495-502.

22. Soria JM, Martínez Ramos C, Bahamonde O, García Cruz DM, Salmerón Sánchez

M, Casas C, et al. Influence of the substrate’s hydrophilicity on the in vitro

Schwann cells viability. J Biomed Mater Res Part A 2007;83:463-470.

27

23. Martínez-Ramos C, Lainez S, Sancho-Bielsa F, García Esparza A, Planells-Cases R,

García Verdugo JM, et al. Differentiation of postnatal neural stem cells into glia and

functional neurons on polymeric substrates. Tissue Eng 2008;14:1365-1375.

24. Martínez-Ramos C, Vallés-Lluch A, García Verdugo JM, Gómez Ribelles JL,

Barcia Albacar JA, Baiget Orts A, et al. Channeled scaffolds implanted in adult rat

brain. J Biomed Mater Res A 2012;100:3276-3286.

25. Rico P, Rodríguez Hernández JC, Moratal D, Monleón Pradas M, Salmerón Sánchez

M. Substrate-induced assembly of fibronectin into networks. Influence of surface

chemistry and effect on osteoblast adhesion. Tissue Eng 2009;15:3271-3281.

26. Soria JM, Sancho-Tello M, García Esparza MA, Mirabet V, Bagan JV, Monleón M,

et al. Biomaterials coated by dental pulp cells as substrate for neural stem cell

differentiation. J Biomed Mater Res Part A 2011;97:85-92.

27. Ying M, Saha K, Bogatyrev SR, Yang J, Hook AL, Kalcioglu ZI, et al.

Combinatorial development of biomaterials for clonal growth of human pluripotent

stem cells. Nature Materials 2010;9:768–778.

28. Campillo Fernández AJ, Arredondo Zamudio AA, Gallego Ferrer G, Monleón

Pradas M, Carda Batalla MC. Polymeric scaffolds for guided dentine regeneration.

Int J Artif Organs 2005;28:362.

29. Brígido Diego R, Pérez Olmedilla M, Serrano Aroca A, Gómez Ribelles JL,

Monleón Pradas M, Gallego Ferrer G, et al. Acrylic scaffolds with interconnected

spherical pores and controlled hydrophilicity for tissue engineering. J Mater Sci:

Mater Med 2005;16:693-698.

30. Más Estellés J, Krakovsky I, Rodríguez Hernández JC, Piotrowska AM, Monleón

Pradas M. Mechanical properties of porous crosslinked poly(ethyl-acrylate) for

tissue engineering. J Mater Sci 2007;42:8629-8635.

28

31. Rodríguez Hernández JC, Serrano Aroca A, Gómez Ribelles JL, Monleón Pradas M.

Three-dimensional nanocomposite scaffolds with ordered cylindrical orthogonal

pores. J Biomed Mater Res B Appl Biomater 2008;84:541-549.

32. Shafy A, Fink T, Zachar V, Lila N, Carpentier A, Chachques JC. Development of

cardiac support bioprostheses for ventricular restoration and myocardial

regeneration. Eur J Cardiothorac Surg 2013;43:1211-1219.

33. Detournay E, Cheng AH-D. Fundamentals of poroelasticity. In: Fairhurst C, editor.

Comprehensive rock engineering: principles, practice and projects, vol. II, Analysis

and design method. New York: Pergamon Press, 1993. pp. 113-171.

34. Fornells P, García-Aznar JM, Doblaré M. A finite element dual porosity approach to

model deformation-induced fluid flow in cortical bone. Ann Biomed Eng

2007;35:1687-1698.

35. Jiang W, Schwendeman S. Stabilization and controlled release of bovine serum

albumin encapsulated in poly(D, L-lactide) and poly(ethylene glycol) microsphere

blends. Pharmaceut Res 2001;18:878-885.

36. Khoury C, Adalsteinsson T, Johnson B, Crone WC, Beebe DJ. Tunable

microfabricated hydrogels – A study in protein interaction and diffusion. Biomed

Microdev 2003;5:35-45.

37. Ritger PL, Peppas NA. A simple equation for description of solute release. I. Fickian

and non-Fickian release from non-swellable devices in the form of slabs, spheres,

cylinders or discs. J Controlled Release 1987;5:23–36.

38. Naddaf AA, Tsibranska I, Bart HJ. Kinetics of BSA release from poly(N-

isopropylacrylamide) hydrogels. Chem Eng Proc 2010;49:581-588.

39. Lin C-C, Metters AT. Hydrogels in controlled release formulations: network design

and mathematical modeling. Adv Drug Deliv Rev 2006;58:1379-1408.

29

40. Masaro L, Zhu XX. Physical models of diffusion for polymer solutions, gels and

solids. Prog Polym Sci 1999;24:731-775.

41. He H. Multifunctional medical devices based on pH-sensitive hydrogels for

controlled drug delivery. PhD dissertation, The Ohio State University, 2006.

42. Jiang W, Schwendeman S. Stabilization and controlled release of bovine serum

albumin encapsulated in poly(D, L-lactide) and poly(ethylene glycol) microsphere

blends. Pharmaceut Res 2001;18:878-885.

43. Huang X, Brazel CS. On the importance and mechanisms of burst release in matrix

controlled drug delivery sustems. J Controlled Release 2001;73:121-136.

30

Figure captions

Figure 1. SEM images of the PEA scaffolds. (a) cross section, (b) surface.

Figure 2. CryoSEM cross section images of: (a) PEA scaffold loaded with PBS, PEA scaffolds loaded with

0.15% SAP solution (b) and after gelling with PBS (c, d at different magnification). Insets in (a, c) are

macroscopic images of PEA scaffolds after congo red staining: (a) bare and (c) loaded with the SAP gel

Figure 3. Stress-strain curves obtained from compressive experiments performed on PEA scaffolds: bare,

and filled with the SAP gel. Insets: initial Young modulus, E1, and linear modulus E2 calculated at strain

0.3.

Figure 4. BSA release curve from loaded scaffold-cum-gel constructs in PBS: fractional mass Mt/M0 of

BSA released at time t versus time t. Inset: fit to a power law function of the data of the initial release

regime.

Figure 5. Fluorescence microscopy images of the surfaces of PEA scaffolds (without and with SAP gel

within their pores) seeded with fibroblasts (statically and dynamically) and cultured for 1 day (scale bar 200

μm) and 7 days (scale bar 100 μm); DAPI (nuclei) staining in blue.

Figure 6. Fluorescence microscope images of fibroblasts seeded statically and dynamically in PEA

scaffolds with and without SAP solution in their pores, cultured for 7 days. DAPI stain for nuclei (blue),

phalloidin stain for actin (green) and vimentin stain for cell membrane (red). Images correspond to 100 μm

thick internal slices.

Figure 7. Confocal laser microscope images of fibroblasts seeded dynamically in PEA scaffolds with the

SAP gel in their pores, cultured for 1 and 7 days. DAPI stain for nuclei (blue) and phalloidin stain for actin

(green). Images correspond to 100 μm-thick internal slices.

Figure 8. SEM images at different magnifications of cross sections of PEA scaffolds with SAP gel in their

pores, seeded dynamically with fibroblasts and cultured for 7 days.

Figure 9. Sheep adipose-tissue derived stem cells (ASCs) seeded dynamically in a PEA scaffold filled with

the SAP and cultured for 7 days: (a) CLSM image of a 100 μm-thick internal slice, DAPI stains for nuclei

(blue) and phalloidin stains for actin (green); (b) SEM image of the surface.

Graphical abstract

(a) SEM image of a PEA scaffold with spherical interconnected pores. (b) CryoSEM cross section image of

a PEA scaffold loaded with 0.15% (w/v) SAP solution and gelled with PBS. (c) CLSM image of fibroblasts

seeded dynamically in PEA scaffolds with the SAP gel in their pores, cultured for 7 days. DAPI stain for

nuclei (blue) and phalloidin stain for actin (green). Image corresponds to a 100 μm thick internal slice.

Graphical Abstract (for review)

Figure 1Click here to download high resolution image

Figure 2Click here to download high resolution image

Figure 3Click here to download high resolution image

Figure 4Click here to download high resolution image

Figure 5Click here to download high resolution image

Figure 6Click here to download high resolution image

Figure 7Click here to download high resolution image

Figure 8Click here to download high resolution image

Figure 9Click here to download high resolution image